A novel cumulative level difference mean based GLDM and modified ABCD features ranked using eigenvector centrality approach for four skin lesion types classification

• Published in: Computer methods and programs in biomedicine Vol. 165; pp. 163 - 174
• Main Authors: Wahba, Maram A., Ashour, Amira S., Guo, Yanhui, Napoleon, Sameh A., Elnaby, Mustafa M. Abd
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01.10.2018
• Subjects: Algorithms; Carcinoma, Basal Cell - classification; Carcinoma, Basal Cell - diagnostic imaging; Carcinoma, Basal Cell - pathology; Carcinoma, Squamous Cell - classification; Carcinoma, Squamous Cell - diagnostic imaging; Carcinoma, Squamous Cell - pathology; Cumulative level-difference mean; Databases, Factual; Dermoscopy - methods; Diagnosis, Computer-Assisted - methods; Diagnosis, Computer-Assisted - statistics & numerical data; Diagnosis, Differential; Feature ranking; Fractals; Humans; Image Interpretation, Computer-Assisted - methods; Image Interpretation, Computer-Assisted - statistics & numerical data; Keratosis - classification; Keratosis - diagnostic imaging; Keratosis - pathology; Melanoma - classification; Melanoma - diagnostic imaging; Melanoma - pathology; Modified-ABCD feature vector; Nevus, Pigmented - classification; Nevus, Pigmented - diagnostic imaging; Nevus, Pigmented - pathology; Pattern Recognition, Automated - methods; Pattern Recognition, Automated - statistics & numerical data; Skin - diagnostic imaging; Skin - pathology; Skin Diseases - classification; Skin Diseases - diagnostic imaging; Skin Diseases - pathology; Skin lesion classification; Skin Neoplasms - classification; Skin Neoplasms - diagnostic imaging; Skin Neoplasms - pathology; Support Vector Machine; Skin lesion classification; Support vector machine; Feature ranking; Cumulative level-difference mean; Modified-ABCD feature vector
• ISSN: 0169-2607; 1872-7565; 1872-7565
• DOI: 10.1016/j.cmpb.2018.08.009

•This study proposes an automated skin lesion detection system as an early warning tool for skin lesion classification.•The proposed method combined the a new texture feature (CLDM) based on texture GLDM which is combined with the most discriminative modified-ABCD features as determined by Eigenvector Centrality (ECFS) ranking algorithm to classify the targeted classes.•These extracted features achieved outstanding performance in terms of all metrics as accuracy, sensitivity, specificity of 100% using Q-SVM. Melanoma is one of the major death causes while basal cell carcinoma (BCC) is the utmost incident skin lesion type. At their early stages, medical experts may be confused between both types with benign nevus and pigmented benign keratoses (BKL). This inspired the current study to develop an accurate automated, user-friendly skin lesion identification system. The current work targets a novel discrimination technique of four pre-mentioned skin lesion classes. A novel proposed texture feature, named cumulative level-difference mean (CLDM) based on the gray-level difference method (GLDM) is extracted. The asymmetry, border irregularity, color variation and diameter are summed up as the ABCD rule feature vector is originally used to classify the melanoma from benign lesions. The proposed method improved the ABCD rule to also classify BCC and BKL by using the proposed modified-ABCD feature vector. In the modified set of ABCD features, each border feature, such as compact index, fractal dimension, and edge abruptness is considered a separate feature. Then, the composite feature vector having the pre-mentioned features is ranked using the Eigenvector Centrality (ECFS) feature ranking method. The ranked features are then classified by a cubic support vector machine for different numbers of selected features. The proposed CLDM texture features combined with the ranked ABCD features achieved outstanding performance to classify the four targeted classes (melanoma, BCC, nevi and BKL). The results report 100% outstanding performance of the sensitivity, accuracy and specificity per each class compared to other features when using the highest seven ranked features. The proposed system established that Melanoma, BCC, nevus and BKL are efficiently classified using cubic SVM with the new feature set. In addition, the comparative studies proved the superiority of the cubic SVM to classify the four classes.